您可以直接在 UI 中為特定資料類型建立 AutoML Edge (可匯出的) 模型,也可以以程式輔助方式啟動訓練管道工作。您可以使用準備好的資料集建立這個模型。您可以在 Google Cloud 控制台中建立這個資料集,也可以使用 API 建立。Vertex AI API 使用資料集項目訓練模型、測試模型,並評估模型效能。查看評估結果,視需要調整訓練資料集,並使用改善的資料集建立新的訓練工作。
訓練工作可能需要數小時才能完成。 Google Cloud 控制台的 Vertex AI 頁面會顯示訓練狀態。
[[["容易理解","easyToUnderstand","thumb-up"],["確實解決了我的問題","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["難以理解","hardToUnderstand","thumb-down"],["資訊或程式碼範例有誤","incorrectInformationOrSampleCode","thumb-down"],["缺少我需要的資訊/範例","missingTheInformationSamplesINeed","thumb-down"],["翻譯問題","translationIssue","thumb-down"],["其他","otherDown","thumb-down"]],["上次更新時間:2025-09-04 (世界標準時間)。"],[],[],null,["# Train an AutoML Edge model using the Google Cloud console\n\nYou create an AutoML Edge (exportable) model directly in the UI for certain\ndata types, or by starting a training pipeline job\n[programmatically](/vertex-ai/docs/training/automl-edge-api). You create this model using a prepared\ndataset. Create this dataset in the Google Cloud console or\nusing the [API](/vertex-ai/docs/training/automl-edge-api). Vertex AI API uses the\nitems from the dataset to train the model, test it, and evaluate\nmodel performance. Review the evaluations results, adjust the training dataset\nas needed, and create a new training job using the improved dataset.\n\nTraining jobs can take several hours to complete. The Vertex AI\npage of the Google Cloud console shows the status of training.\n\nTraining an AutoML Edge model\n-----------------------------\n\n1. In the Google Cloud console, in the Vertex AI section, go to\n the **Datasets** page.\n\n [Go to the Datasets page](https://console.cloud.google.com/vertex-ai/datasets)\n2. Click the name of the dataset you want to use to train your model to open\n its details page.\n\n3. If your data type uses annotation sets, select the annotation set you want\n to use for this model.\n\n4. Click **Train new model**.\n\n5. In the **Train new model** page, complete the\n following steps for your data type:\n\n ### Image\n\n 1.\n Select radio_button_checked**AutoML Edge**\n for the training method and click **Continue**.\n\n 2. Enter the display name for your new model.\n\n 3.\n If you want manually set how your training data is split, expand **Advanced\n options** and select a data split option.\n [Learn more](/vertex-ai/docs/general/ml-use).\n\n 4. Click **Continue**.\n\n 5.\n ***Classification** models only (optional)* : In the **Explainability**\n section, select check_box**Generate explainable\n bitmaps for each image in the test set** to enable\n [Vertex Explainable AI](/vertex-ai/docs/explainable-ai/overview).\n Choose [visualization settings](/vertex-ai/docs/explainable-ai/visualization-settings-automl-icn) and\n click **Continue**.\n\n This feature has costs associated with it. See [Pricing](/vertex-ai/pricing)\n for more information.\n\n \u003cbr /\u003e\n\n 6.\n Select the optimization goal that best suits your need. You\n can optimize for accuracy, latency, or both.\n\n 7. Click **Continue**.\n\n 8.\n In the **Compute and pricing** window, enter the maximum number of\n hours you want your model to train for.\n\n\n This setting helps you put a cap on the training costs. The actual\n time elapsed can be longer than this value, because there are other\n operations involved in creating a new model.\n\n \u003cbr /\u003e\n\n 9.\n If you want to stop training when the model is no longer\n improving, select **Enable early stopping**.\n\n ### Video\n\n 1. Enter the display name for your new model.\n\n 2. Click **Continue**.\n\n 3.\n Select radio_button_checked**AutoML Edge**\n for the training method and click **Continue**.\n\n 4.\n Select the optimization goal that best suits your need. You\n can optimize for accuracy, latency, or both.\n\n 5. Click **Continue**.\n\n Several minutes after training starts, you can check the training\n node hour estimation from the model's properties information.\n If you cancel the training, there is no charge on the current product.\n\n \u003cbr /\u003e\n\n6. Click **Start Training**.\n\n Model training can take many hours, depending on your training budget\n (image only) and the size and complexity of your data. You can close\n this tab and return to it later. You will receive an email when your\n model has completed training.\n\nWhat's next\n-----------\n\n- [Evaluate AutoML models](/vertex-ai/docs/training/evaluating-automl-models).\n- [Export AutoML Edge models](/vertex-ai/docs/export/export-edge-model).\n- [Use Vertex Explainable AI to understand model behavior](/vertex-ai/docs/explainable-ai/overview)."]]